Prius Battery rebalance thread

Take what I say with a grain of salt, as I am working on my first cycle, and my setup is just plain not working.

13 is a no brainer. And I can see your concerns on #5. But some of those others like 12 seem to still be improving, I would be tempted to put another 2 cycles on them myself.

I have 2 Gen3 batteries, and I may have some Gen 3 cells for you. First I need to get my setup working before I would be comfortable that they are good , I need to know what the heck I have first.

 

As bad modules go, yours look pretty good! I agree with SEaton that you should give them a few more cycles. The fact that you are hitting your 7250 mAHr charging cutoff means you have not hit the delta-V level of those suspect modules (at least in the last cycles). That means you are not charging the modules to their full capacity. I would suggest putting a few more cycles on those modules ... possibly with a higher charging threshold, e.g. 7500 mAHr.

It appears a little concerning that the capacities seem to have decreased on some of those 4th discharges. I don't know the timing of your process, but on mine I followed the typical script of running a 3-cycle run on every module and then coming back to hit the suspect ones with additional cycles. Having only 3 or 4 chargers (one died), there was quite a time lapse (up to 3 weeks) between my initial cycles and the follow-up. (I am jealous of your effective 8 chargers.) In the interim, there was a nontrivial self-discharge that led to the appearance of worsening capacity. (NiMH has a somewhat high self-discharge immediately after a full charging, e.g. 5-10% over the first 24 hrs. It then settles down to a much slower self-discharge. You don't see this running cycles back-to-back but when the modules are left for several days, it becomes apparent.) It would be more informative on the subsequent iterations to run at least 2 cycles that are back-to-back (or nearly so). You might find that the apparent degradations were erroneous.

In terms of expected module life, it would be helpful to measure the internal battery resistance to see if the suspect modules are out of line. If you cannot measure the internal resistances, the loaded voltages would be a reasonable indirect measure so I suggest you post them.

FYI, I had a goal of getting to at least 6000 mAHr on my modules. I got them to a range of 6.0 to 6.3 AHr. I had to put 9 cycles on the 2 most stubborn modules. (Yes, those modules lost a little lifespan due to this exercising.)

Keep us informed and thanks for posting your data! If you decide to trash those suspect modules, let me know. I would be interested in experimenting on them.

 

On second thought, I don't want to kill modules with 5800-5900 mAHr. I agree there are likely folks out there who would like to use them in a pack. I was planning to do some experiments such as seeing how many cycles the batteries would go (full charges) and comparing against internal resistance. My testing would be fatal so I would prefer to get modules in worse shape than yours.

That is actually a pretty wide variation in loaded voltages but that may not be meaningful given that it is post balance. Some of the modules have presumably self-discharged. Sorry for the follow-up request, but could you post your latest loaded voltage, open-circuit voltage, and loaded current in the same chart for easy comparison? I was not clear before, but it is the difference in the open-circuit and loaded values that is important for each module ... especially when the modules are at differing states of charge.

Regarding battery resistance, it is possible theoretically to derive internal resistance from the open-circuit voltage of the battery and the loaded voltage of the battery if you know the resistance of the load, e.g. see Battery Internal Resistance

However, most light bulbs (including halogen headlamps as I presume you are using) have a resistance that varies with temperature where the temperature varies with the voltage applied. In effect, it is a variable resistor meaning that it's not practical to use R(load) in the equation.

With a variable/unknown load, the method for measuring internal resistance involves loading the battery, measuring the battery voltage, and measuring the current draw through the load. Then, take the open-circuit measurement after the load draw. The internal battery resistance is then:
R(bat)=[V(open_circuit)-V(loaded)]/I(under load)]

You listed the loaded current and voltage so you almost have all the values needed. The only parameter missing is the corresponding open-circuit voltages. Did you take the loaded voltage and current at the same time, i.e. using 2 meters? (Ideally, these would be measured simultaneously using with 2 meters. Taking loaded current and loaded voltage on separate measurements is suboptimal but the values should be close enough in practice.) Google battery internal resistance for more info. FYI, my modules each had internal resistances of around 0.02 ohms each. Britprius and others indicated that was good/typical.

Sorry to hear you haven't been able to use your full charging capability. I just got rid of a bunch of banana jacks and related on ebay so I could have helped you. (They were cheap ... I was just trying to clear out junk.)

I am looking forward to seeing if the repeated cycles help out your modules.
-Marc

 

Sounds like good progress. I hope the added cycles have the desired outcome. Go module 12!

-Marc

 

Wow, that capacity increased nicely.

I understand your concern about the self-discharge (18% over 10 days by my arithmetic). I don't have enough experience with the Prius modules to know if that is typical. A typical NiMH cell discharges 5-10% in the first 24 hours after a full charge and settles down to on the order of 1% per day thereafter. However, I believe these Prius cells are supposed to exhibit better self-discharge characteristics. There is a somewhat recent thread on the board regarding self-discharge during battery storage. Search PriusChat for self discharge and see if you can find that thread. (JeffD had some good comments.) Or maybe some experienced member has some data for self discharge rates they can share here?

I wouldn't worry about not hitting the delta V cutoff. That module is now on the strong side (relative to your other modules) so I don't think you want it to recover more capacity (or heat that module more than necessary). It is interesting that your first 2 cycles did hit the delta V. This is a good example of recovering capacity. The increased charging capacity (7500 mAHr) I suggested was probably not even necessary.

Let us know the results from your other suspect modules. Based on the module 12 results, I presume you'll be hitting the others with more cycles.

 

Is there a good idea as to what exactly is happening in the battery when it gets deep cycled, such that its capacity will increase? I know about lead acid batteries, and the way in which deep cycles will break up sulfation and in general create slightly more surface area in the plates. How does it work in the case of Prius batteries?

Or is it more about equalization of the series strings?

 

Great job, I am jealous of those numbers. I have 2 Gen III's and a Gen I I am working on right now. Its a long story, but I thought I was nuts and had a real bad setup nothing made sense, then I used the same setup on the other Gen III and wow, tons better numbers.

So what speed are you charging at ? just curious, I can charge anywhere from real slow to 5A with my setup. I have chargers I don't think anyone else has used these racers edge ones. that was probably a mistake. I don't have delta V, they have delta peak. but anyhow my numbers are finally getting realistic looking.

Glad you redid those last cycles, please keep us posted on your progress as you go

 

It is not that the capacity of the modules has increased. It is just one or more cells within the module has for some reason become more discharged than the others. Charging the module to 7500 MaH over charges the full cells while allowing the those with a lower state of charge to catch up.
Imagine a situation where one cell has a charge level 2500 Mah bellow the rest. Remember the capacity of a module is that of it's lowest cell.
Charging the module from fully discharged to 6500 Mah "the theoretical module capacity" would leave that low cell in the same state of charge compared to the rest.
Charging to 7500 Mah would over charge the full cells by 1000 MaH, giving off the overcharge as heat. However the low cell will continue to charge and catch up by 1000 MaH. Meaning it is now only 1500 MaH below the rest.
Cycling again would add another 1000 MaH, and again would bring it up to the level of the rest of the cells and be slightly over charged.
This is a simplistic view because I have not taken into account things like charging losses, and non linear charging characteristics, but it does give you the basis of what balancing does.

John (Britprius)

 

OK John, thanks. I guess the question is about that "for some reason" aspect that makes one cell fall out of sync. Over the life of a typical battery, there have been thousands of small, and fewer large, charge-discharge cycles. That the cells have stayed in sync for such a period of use would say to me that there is either a miraculous level of manufacturing precision that makes for excellent cell balancing natively, or there must be some mechanism, perhaps in the battery controller, that does some sort of equalization on a periodic basis. But the latter aspect is perhaps not possible since some Prius' may never see all green bars in their life.

While I am glad that people are able to get more life out of their used batteries, and may have to do it myself someday, I really do wonder what the cycle limit of these cells are. It is too soon to tell if people are going to end up playing "whack-a-mole."

 

Thanks for the information. I have only been charging mine to 7200 or 7300 mah, maybe way too low. My charges always hit the capacity limit. I have been afraid to let one charge till it hits delta peak limit.

Let me know how it goes when you have it all back together, how things go

 

Great thread and data on balancing the modules within the battery and I've got basically the same setup ready to go myself (using two Hitec X4 AC Plus chargers). Just an FYI for others since it seems like the Thunder T6 is identical to what I'm using.

I use the following:
Mode: Program Save then use Program Load when cycling other modules
Battery type: NiMH
Nominal voltage: 7.2V (9.6V on the Highlander)
Charge current: 2A (2A-5A, lower is better for heat buildup in modules)
Discharge voltage: 6V (8V on the Highlander)
Discharge current: 1A (use highest setting available)
Charge capacity: 7000 mAh (range 6500-7500 mAh, higher capacity should use lower charge currents)
DCHG>CHG: 3 (more cycles if needed)

Also go into User Set Program and set:
Capacity Cutoff: on, 7000 mAh
Safety Time: off, (time doesnt matter)
Temp Cutoff: 45C (if you are using temp probes)
Waste Time: 5mins

After that then run your cycles and keep track of the capacity improvements.

Once the cells have been exercised and all show proper capacity they should be brought back down to normal operating voltage range (ie between 40-80% of battery capacity) or the vehicle will throw an error code. So for a Prius you'd want to aim for around 7.6-8.0V (or around 10.4-10.8V for the Highlander). Then connect all the positives and all the negatives (ie creating one giant 7.2V battery) to let all the voltages equalize among modules (usually overnight). Once that is done and the voltage drift is acceptable then go ahead and reassemble the battery pack for re-installation.

 

If any one has settings for the Supermate DC6 I would be grateful.

 

One more note is to make sure to use good quality cables (10A min rating) when cycling your modules. I got some cheaper banana test cables and it wouldnt charge properly when I did some tests. So I went and made my own using high quality speaker cable (monster cable) and it worked much better after that. You can buy good quality banana terminals on ebay for fairly cheap and making your own cable lengths will allow you to tailor your own setup perfectly.

I edited my settings above to what I am currently using.